HumanG_Fusion: RNA-Seq Fusion Detection Pipeline

HumanG_Fusion is a modular, Nextflow-based pipeline for human RNA-Seq fusion detection, featuring:

• Preprocessing (FastQC, trimming)
• Optional Kraken2 decontamination
• STAR alignment (2-pass)
• Fusion calling (STAR-Fusion, Arriba, FusionCatcher)
• Postprocessing & merging with gene annotation

---

📦 Repository Layout

HumanG_Fusion/
├── README.md
├── installation_setup/
│   ├── Docker_setup/
│   ├── bash_setup/
│   └── python_setup/
├── docker/
├── bin/
├── conf/
├── modules/
├── scripts/
├── bash_scripts/
├── tools/
├── sample_data/
├── main.nf
├── pipeline.config.sh
└── run_pipeline.sh

---

🚀 Getting Started

1️⃣ Installation Setup

Choose one of the following methods:

---

🐳 Docker

Best for: Reproducible, containerized execution.

Setup:

cd installation_setup/Docker_setup
# Follow README.md for build/run instructions
docker build -t humang_fusion:latest .

Or, use Docker Compose:

cd docker_compose_setup
docker compose build

---

📦 Bash

Best for: Quick setup on Linux.

Setup:

cd installation_setup/bash_setup
chmod +x install.sh
./install.sh

This will:

• Install required apt packages (if sudo)
• Install Miniconda under ~/miniconda3 (if missing)
• Create a bioinf environment via mamba
• Install all conda + pip packages

Activate environment:

source ~/miniconda3/etc/profile.d/conda.sh
conda activate bioinf
# or, if fallback venv was used:
source ~/pyenv-bioinf/bin/activate

---

🐍 Python

Best for: Python-centric environments.

Setup:

cd installation_setup/python_setup
chmod +x install.py
./install.py

This will:

• Install Miniconda (if missing)
• Create a bioinf environment
• Install all dependencies

Activate environment:

source ~/miniconda3/etc/profile.d/conda.sh
conda activate bioinf

---

2️⃣ Data Preparation

Place your FASTQ files in sample_data/:

sample_data/
├── sample1_R1.fastq.gz
├── sample1_R2.fastq.gz
├── sample2_R1.fastq.gz
└── sample2_R2.fastq.gz

---

2.1 Reference Preparation

Please Visit ➡️ References

3️⃣ Pipeline Configuration

Edit pipeline.config.sh with your paths:

#!/bin/bash
export READS="./sample_data/*_R1.fastq.gz"
export OUTDIR="./results"
export STAR_INDEX="/refs/STAR_index_GRCh38"
export GENOME_FASTA="/refs/GRCh38.fa"
export GTF="/refs/gencode.v41.annotation.gtf"
export STAR_FUSION_CTAT_LIB="/refs/ctat_resource_lib"
export KRAKEN2_DB="/refs/kraken2_db"
export THREADS=8
export KRAKEN_DECONTAM=true
export RUN_FUSIONCATCHER=false

---

4️⃣ Run the Pipeline

🏃 Automated (Recommended)

chmod +x run_pipeline.sh
./run_pipeline.sh

This script will:

• Source your config
• Run Nextflow with the correct profile (Docker, Conda, or Bash)
• Execute all steps in order

🛠 Manual Step-by-Step

If you prefer, you can run each step manually using the scripts in bash_scripts/ or scripts/.

---

🔧 Pipeline Modules

Module	Description
preprocess	FastQC, trimming (cutadapt)
kraken_decontam	Optional: Remove non-human reads
star_align	STAR 2-pass alignment
fusion_callers	STAR-Fusion, Arriba, FusionCatcher (optional)
postprocess	Merge calls, annotate genes, produce summary TSV

---

📂 Outputs

• results/{sample}/fastq_trimmed/: Trimmed FASTQ
• results/{sample}/align/: BAM files
• results/{sample}/fusion/: Caller-specific outputs
• results/reports/: Merged, annotated fusion calls

---

🧪 Testing & Validation

• Test on small datasets with known fusions.
• Validate artefacts in IGV.

---

📝 Notes

• STAR-Fusion CTAT resource library must be provided.
• Kraken2 DB required if using decontamination.
• mygene is used for gene annotation.

---

🚀 Next Steps

• Add Singularity/Docker push automation
• Add Nextflow Tower profiles
• Create a tests/ folder with test FASTQs

---

💬 Questions?

Open an issue or contact the maintainers!

Justiceoheneamofa@gmail.com / kdanquah@atu.edu.gh / kdanquah@noguchi.ug.edu.gh